CV Dazzle - how to hide from face recognition
Written by David Conrad   
Thursday, 04 August 2011

Dazzle camouflage techniques that proved effective in hiding ships in the 20th century might now help you to maintain your anonymity.

The progress in facial recognition algorithms has been such that you can now say that it is a mainstream application. Facebook lets you tag photos automatically and there are lots of websites plus police and security agencies of all types using it. From the algorithmic point of view it is nothing but a success story ... but...

As you can probably imagine, there are people who would rather remain anonymous and untracked and, as with every action, there is a reaction - even if it isn't equal just at the moment.

CV Dazzle is a simple idea. It is camouflage for the face. The name and the overall approach comes from the invention and use of Dazzle camouflage during World War I. The basic idea is not to make the object blend in with the background but to break up its shape by using bright high-contrast shapes - you might be able to see something but not work out what it is.

 

dazzleship

 

Now the same idea has been tried by a student entry  to the Core77 Design awards from Adam Harvey of the  Tisch School of Arts, New York University. The idea is that by augmenting the face and occasionally hiding some aspects of it, face recognition software just can't do its job. As well as just coming up with effective dazzle designs it is also important to find something that is wearable and enhances the overall look and style of the person using it. In simple terms it has to be feasible that someone could wear CV Dazzle in a normal situation!

Some of the designs involve using the hair style to obscure parts of the face - see CV Dazzle. Notice that wearing masks in public is illegal in some countries, but CV Dazzle probably doesn't fall into the category of "masks".

 

dazzlegirl

 

The designs have only been tested against OpenCV and Facebook and which means that there is no guarantee that they will work against a security grade face recognition system. However, if you know how face recognition works it seems reasonable that dazzle camouflage would at least reduce the identification certainty.

Finally the creator of CV Dazzle offers some tips on staying unrecognized:

  1. Avoid enhancers: They amplify key facial features.
  2. Partially obscure the nose-bridge. The region where the nose, eyes, and forehead intersect is a key facial feature.
  3. Partially obscure the ocular region. The position and darkness of eyes is a key facial feature.
  4. Remain inconspicuous. For camouflage to function, it must not be perceived as a mask or disguise.

So what is the next reaction?

Better facial recognition software seems the only option and given that CV Dazzle doesn't seem to interfere much with a human's ability to recognize the wearer it should still be possible.

 

If you would like to be informed about new articles on I Programmer you can either follow us on Twitter or Facebook or you can subscribe to our weekly newsletter.

Banner


Microsoft Introduces Vector Abstractions Library For .NET
21/11/2024

Microsoft has announced a preview release of the Microsoft Extensions VectorData Abstractions library, which can be used to help integrate vector stores into .NET applications and libraries.



IBM Updates Granite Models
28/10/2024

IBM has released new Granite models that it says provide state-of-the-art performance relative to model size. The Granite 3.0 collection includes a new, instruction-tuned, dense decoder-only LLM.


More News

Last Updated ( Thursday, 04 August 2011 )